Deployment of Heterogeneous UAV Base Stations for Optimal Quality of Coverage

This paper studies the quality of coverage of deploying flying base stations mounted on unmanned aerial vehicles (UAV-BSs) after disasters or during some occasional events. In particular, we focus on the problem of minimizing the average UAV-user distance, while maintaining connectivity between the UAV-BSs and some nearby stationary base stations (SBSs). The UAV-BSs can be deployed at different altitudes, and their transmission powers may also be different. We first propose a decentralized deployment algorithm for a line-of-sight (LoS) scenario. This algorithm allows UAV-BSs to determine their movements based on only local information. So, it is applicable in large-scale. The local optimality and the convergence of the algorithm are proved. Moreover, we discuss how to use the algorithm in non-line-of-sight (NLoS) scenarios. Specifically, during its movement, each UAV-BS needs to verify the connectivity requirement as well as if a future movement will lose any already covered users. This extension guarantees that the average UAV-user distance keeps reducing during the movements of UAV-BSs. Computer simulations and comparison with a benchmark method confirm the effectiveness of the proposed algorithms in terms of the quality of coverage.